Sandia National Labs SA3000 8085 CPU

The CPUShack Museum writes that back in the late 1970’s and early 1980’s Sandia National Laboratory (in Albuquerque, New Mexico) began building the capacity to design, fab, and test IC’s at scale (packaging was handled by Fairchild and Allied Signal).

Why would a National Laboratory need the capacity to do this? To provide components that were not available commercially.  In this case, Sandia’s goal was to make radiation hardened devices for use in weapons and space missions.  The harshest environments and where reliability was paramount.

In 1982 Sandia began work on a CMOS rad-hard conversion of the Intel 8085 processor.  This would become the Sandia SA3000.

Converting the HMOS Intel 8085 to a rad hard CMOS process took some doing.  The original 8085 had around 6500 transistors.  Converting it to CMOS resulted in a 18,000 transistor device. One of the trickier conversions was the instruction decoder (a large PLA), easy in NMOS, less so in CMOS. The SA3000 was made on 4″ wafers on a 3u process.

In around 1984-1985 the govt decided to bring in a contractor to run the fab much to the annoyance of Sandia management and detriment to efficiency.  This would be Allied Signal, who at that point had no experience running fab’s.  This greatly slowed down production for some time.

The SA3000 (and its support chips) were commercialized by Harris in 1990 as the HS1-80C85RH and the HS9-80C85RH.  These are similar to the SA3000, made on the same process, but spec’d with only a 5V operating voltage rather than 10V and a max speed of 2MHz rather then the 10MHz the SA3000 is capable of.  The HS1 is space grade, with higher levels of screening, and the HS9 is military grade, but not screened to space use levels.

See much more in the article here.

Intel 8085A die